With the advent of healthcare reform, it's more important than ever for healthcare providers to identify and target high-risk patients to prevent readmissions. The problem is determining which model best predicts those patients.
Hospitals can rely on several specific models to help coordinate care, according to research published in the American Journal of Managed Care.
Lead researcher Lindsey R. Haas of the Mayo Clinic and her team studied seven risk-adjustment models and their effectiveness in predicting hospitalizations, 10-day readmissions, high expenditures and emergency room visits. The models they evaluated were:
Adjusted Clinical Groups, which are based on the absence or presence of specific diagnoses from inpatient and outpatient services over a specific period of time;
Minnesota Tiering, which organizes patients into "complexity tiers" based on the number of major condition categories they fall under;
Hierarchical Condition Categories, a model implemented by the Centers for Medicare & Medicaid Services, which involves the aggregation of diagnosis codes and demographic data into 70 condition categories to generate a single risk score;
Elder Risk Assessment (ERA) Index, which generates a risk score based on a combination of demographic information, selected medical conditions and number of hospital days in the last two years;
Charlson Comorbidity Index, which sums weights for 17 conditions;
Chronic Condition Count, a comorbidity count that adds up several chronic conditions and groups the sum into one of six categories; and
The Hybrid Model, a combination of Minnesota Tiering and the ERA Index.
Based on a retrospective cohort analysis of 83,000 patient records, the ACG model was a more accurate predictor than the others, according to the researchers. However, they determined all of the models have fair predictive value.
"Use of any of the tools may provide some support for providers and health plans who undertake case management," the researchers wrote. "Focusing care coordination efforts within the medical home on patients likely to benefit most requires appropriate identification of the highest risk, highest utilizing patients."
Risk-adjustment models have had practical problems in the past. A study in February found the model used by Medicare was skewed by regional variations in frequency of doctor visits. As a result, some providers and plans were overpaid and underpaid, FierceHealthcare previously reported.
To learn more:
- here's the research